Verification of path computational performance in functionally distributed transport networking on next-generation network

Traffic in IP networks has increased and services have diversified over the last few years. IP networks need to have scalability and high-quality data transmissions and they must be able to incorporate new functions. For good processing performance and ease of adding functions to the transport stratum on a next-generation network (NGN), we have proposed a functionally distributed transport networking architecture that separates the control-plane processing part (control element, CE) from the forwarding plane processing part (forwarding element, FE) of the router. In this architecture, the CE is mounted on a general-purpose server platform, and one path-control process in the CE consolidates and processes the path computations and the path settings for two or more routers. The functions and performance of the CE are easily upgraded. Moreover, a particular CE computes the path for an entire network by using the single network-integration path-control method that treats each sub-network as one router. This leads to efficient operation of a large-scale network. In this paper, we argue that the issues of our architecture could be bottleneck factors for the path computational performance in a large-scale network and we predict the influence of these factors by using a simulation environment.